You think your medical degree was hard? It seems you probably have not yet sought a London taxi licence. Whoever wants to be a licensed taxi driver in London has to mentally cram a stupendously large amount. 25,000 streets in all their geographical complexity and at least another 20,000 hotels, attractions and other points of contact are part of a test which is considered among the hardest of its kind. Prospective taxi drivers have to study for it on average for three to four years. There isn’t just the one exam, but thereafter several others. About half the candidates fall through.
More elbow grease in the grey matter
Professor Eleanor Maguire of the Wellcome Trust Centre for Neuroimaging at University College London finds this so fascinating that she has already dedicated several neuroimaging studies to the analysis of London’s taxi drivers. In an earlier work, for example, she compared – using a relatively simple approach – the brain volumes of London taxi drivers with those in similar-aged non-taxi drivers. This work showed that the taxi driver, especially with regard to the rear of the hippocampus (so named because of its resemblance to the seahorse) have more grey matter than non-taxidrivers. That fitted in with the new findings in neurophysiology at the time that the hippocampus is – not only in general learning, but especially in learning processes which relate to spatial orientation – attributed a key role.
In a new study, which is soon to be published in the journal Current Biology, Maguire made it a little more challenging. After all, it would be conceivable in principle that people who naturally have a thick posterior hippocampus effectively, as it were, follow their natural calling and are drawn to taxi driving. However this does not seem to be so, because Maguire and her colleagues this time accompanied a group of would-be taxi drivers in London over several years ahead of their “Mega Test”. They carried out pretty MRI scans several times and the subjects also underwent memory tests.
Like the gym: Only a thicker hippocampus resists the test
And now it gets interesting: At the beginning of the learning marathon, the subjects were in memory test results and also in brain structure more or less equally in the same position. Three or four years later, however, everything was different. For those aspiring taxi drivers who had passed the test, the researcher was able to demonstrate a significant increase in volume of the posterior hippocampus, but not in those who had failed. Real learning enlarges the hippocampus, which in turn improves the chance of success when learning the city plan, seems to now be the postulated causal chain. “The human brain remains malleable even in adult life,” says Maguire. “We have seen directly inside individuals how the hippocampus structure changes due to external stimulation.” Whether the observed changes are due to an increase in nerve cells, or are “only” due to an increase in the links between them, is so far not yet clear. But the fact is that the hippocampus belongs to one of those brain areas about which it is known that wholly new neurons can actually be produced.
It’s not the volume, but the shape that counts
Where there is light, there is always shadow. The hippocampus grows, but it also shrinks, or can change its shape in other ways. This can in certain situations be diagnostically helpful. At the annual meeting of U.S. radiologists (RSNA 2011) in Chicago, the Dutch scientist Hakim Achterberg from the Department of Radiology at the Erasmus Medical Centre in Rotterdam presented the outcomes of a subpopulation of the Rotterdam Scan Study. Achterberg reported on 511 subjects who initially received an MRI scan and then, in view of the emergence of dementia, were followed for ten years, by means of cognitive tests and an assessment of general health records. One in ten of these subjects over the course of ten years developed clinical dementia. In order to see to what extent the hippocampus could predict the emergence of dementia, Achterberg and colleagues meticulously segmented this brain region and defined a total of 1024 “milestones markers”, spanning the specific brain segments. From the form of the segments of identical volume, a complex, multidimensional vectorial sum was able to be calculated. And on the basis of this vector an algorithm was then applied in order to predict the emergence of dementia.
Ultimately the scientists could show that, solely based on the shape of the hippocampal formation, dementia was able to be predicted about three to four years before clinical diagnosis. The prediction by virtue of the shape is thus more accurate than previously discussed predictive models which are based solely on hippocampal volume. Further back in time the effect is more blurred. What’s interesting is that there is no form change which would be typical of the emergence of dementia. The human eye no longer helps at this point. Only when the form data is integrated mathematically is a distinction made possible.